Biogeochemical Response of Multiple Iron Redox

1 downloads 0 Views 6MB Size Report
Feb 16, 2017 - BIOGEOCHEMICAL RESPONSE TO MULTIPLE IRON REDOX OSCILLATIONS: ... OSCILLATIONS: LABORATORY AND FIELD INVESTIGATIONS ...... greatly increases the probability that key system components are ...... comprising a 7 d reduction period followed by 7 d of oxidation were employed to.
Biogeochemical Response of Multiple Iron Redox Oscillations: Laboratory and Field Investigations

Item type

text; Electronic Dissertation

Authors

Thompson, Aaron

Publisher

The University of Arizona.

Rights

Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Downloaded

16-Feb-2017 20:56:40

Link to item

http://hdl.handle.net/10150/194955

BIOGEOCHEMICAL RESPONSE TO MULTIPLE IRON REDOX OSCILLATIONS: LABORATORY AND FIELD INVESTIGATIONS

by Aaron Thompson _____________________ Copyright © Aaron Thompson 2005 A Dissertation Submitted to the Faculty of the DEPARTMENT OF SOIL, WATER AND ENVIRONMENTAL SCIENCE In Partial Fulfillment of the Requirements For the Degree of DOCTOR OF PHILOSOPHY In the Graduate College THE UNIVERSITY OF ARIZONA

2005

2

THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE As members of the Dissertation Committee, we certify that we have read the dissertation prepared by

AARON THOMPSON

entitled

BIOGEOCHEMICAL RESPONSE OF MULTIPLE IRON REDOX OSCILLATIONS: LABORATORY AND FIELD INVESTIGATIONS

and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of

Doctor of Philosophy 10/28/2005

_______________________________________________________________________

Jonathan Chorover

____________________

date 10/28/2005

_______________________________________________________________________

Martha Conklin

____________________

date 10/28/2005

_______________________________________________________________________

Jim Field

____________________

date 10/28/2005

_______________________________________________________________________

Raina Maier

____________________

date 10/28/2005

_______________________________________________________________________

Joaquin Ruiz

____________________

date

Final approval and acceptance of this dissertation is contingent upon the candidate’s submission of the final copies of the dissertation to the Graduate College. I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement. 10/28/2005 _______________________________________________________________________

Dissertation Director: Jonathan Chorover

____________________

date

3

STATEMENT BY AUTHOR This thesis has been submitted in partial fulfillment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library. Brief quotations from this thesis are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the copyright holder.

Aaron Thompson

SIGNED: _____________________________

4

ACKNOWLEDGEMENTS I wish to extend gratitude towards and acknowledge the kind assistance of the following people:

Jon Chorover: For his never-ending support of my scientific interests, even those that diverged from his own research program. I have heard Jon described as a breath of fresh air and perhaps that is why I always leave his office more inspired, more knowledgable, and less worried than when I entered. Oliver Chadwick: For opening my mind to how cool soil development and pedology really is. Marykay Amistadi: There is really no way to describe her value to the workings of our laboratory, except essential and irreplaceable. Thank you, thank you, thank you, Marykay. Mark Baker: For his expertise in ICP-MS operation and willingness to help a student— with no knowledge of isotope measurements and who was from an outside department—to develop a complicated new technique on an expensive machine. Committee members: Drs. Martha Conklin, Jim Field, Raina Maier and Joaquin Ruiz for their invaluable scientific counsel. Julie Nelson: For allowing me space, equipment and assistance for my microbial studies. Jennifer Jo Thompson: For our close partnership that has opened so many possibilities for both of us. Without that, none of this could have ever begun. Margaret Helen Thompson: For joyously taking me as far away from academia as possible, every single night.

5

DEDICATION I dedicate this dissertation to:

My parents, Don and Joan Thompson. They taught me how fun and exciting the world could be to ask questions. I’m still asking questions. Thanks.

6

TABLE OF CONTENTS

LIST OF FIGURES.....................................................................................................15

ABSTRACT .................................................................................................................16

CHAPTER 1: INTRODUCTION ...............................................................................18 1. EXPLANATION OF DISSERTATION FORMAT ...................................................18 2. EXPLANATION OF PROBLEM ..........................................................................20 2.1 Research Focus ...............................................................................23 3. CONCEPTUAL FRAMEWORK AND RESEARCH APPROACH ..............................25

CHAPTER 2: PRESENT STUDY ..............................................................................31 1. SUMMARY ......................................................................................................31 1.1 Objective #1....................................................................................31 1.2 Objective #2....................................................................................32 1.3 Objective #3....................................................................................34 1.4 Objective #4....................................................................................35 1.5 Objective #5....................................................................................36

CHAPTER 3: CONCLUSIONS..................................................................................38

REFERENCES ............................................................................................................40

7

TABLE OF CONTENTS – Continued APPENDIX A: IRON-OXIDE CRYSTALLINITY INCREASES DURING SOIL REDOX OSCILLATIONS................................................................................44 ABSTRACT .........................................................................................................45 1. INTRODUCTION ..............................................................................................46 2. MATERIALS AND METHODS ...........................................................................48 2.1 Overall approach ............................................................................48 2.2 Soil selection....................................................................................49 2.3 Initial soil characterization.............................................................49 2.5 Redox oscillation experiment .........................................................52 2.5.1 Reactor operation ....................................................................52 2.5.2 Reactor sampling scheme.........................................................53 2.6 Chemical analysis ...........................................................................54 2.7 Mössbauer Spectroscopy ................................................................56 2.8 Calculations.....................................................................................58 3. RESULTS ........................................................................................................59 3.1 Initial soil characterization.............................................................59 3.2 Iron dynamics .................................................................................61 3.3 57Fe Mössbauer characterization ...................................................67 3.3.1 Soil Fe mineral composition.....................................................68 3.3.2 Changes in Fe mineral composition resulting from redox oscillations .......................................................................................73

8

TABLE OF CONTENTS – Continued 3.4 Metal partitioning...........................................................................75 4. DISCUSSION ...................................................................................................78 4.1 Solid phase Fe speciation ................................................................78 4.2 Iron redox cycles as a biogeochemical forcing function ................80 4.3 Fe redox cycle evaluation................................................................80 4.3.1 Eh, pH and FeIIaq considerations..............................................80 4.3.2 Solid phase Fe distribution and mineral ripening .....................83 4.3.3 Solubility product analysis .......................................................86 4.3.4 Fe reduction/oxidation rates ....................................................88 4.3.5 Solid phase behavior of Al, Si and Ti........................................89 5. CONCLUSIONS................................................................................................90 6. REFERENCES .................................................................................................92

APPENDIX B: MOBILIZATION OF COLLOIDAL REFRACTORY ELEMENTS DURING IRON REDOX OSCILLATIONS ......................................101 1. INTRODUCTION ............................................................................................103 2. MATERIALS AND METHODS..........................................................................106 2.1 Overall approach ..........................................................................106 2.2 Soil Selection .................................................................................106 2.3 Redox reactor design ...................................................................106 2.4 Redox oscillation experiment .......................................................107

9

TABLE OF CONTENTS – Continued 2.4.1 Reactor operation ..................................................................107 2.4.2 Redox sampling scheme .........................................................108 2.5 Batch pH shift experiment............................................................109 2.6 Chemical analysis .........................................................................109 2.7 TEM preparation and analysis.....................................................111 2.8 Notation.........................................................................................112 3. RESULTS ......................................................................................................112 3.1 Colloid chemistry ..........................................................................112 3.2 Batch pH shift Experiment...........................................................115 3.3 TEM/EDS observations ................................................................116 3.4 Redox Oscillation Experiments ....................................................119 3.4.1 Operational parameters: [FeII]10 m2 g-1) and reactive site densities, colloidal particles (ca. 1 nm – 1 µm in size) often regulate the mobility of strongly sorbing constituents in soils, including heavy metals and refractory elements that are otherwise assumed to be “immobile” within the soil profile because of low solubility (Kretzschmar et al., 1999). The mobilization of soil colloids, also termed colloidal dispersion, is strongly dependent on solution pH, ionic strength and ionic composition because these parameters affect the magnitude and sign of particle surface charge and, as a result, interparticle electrostatic repulsion (Suarez et al., 1984; Heil and Sposito, 1993; Bunn et al., 2002; Sposito, 2004). In general, dispersion of soil colloids is favored at pH values above the point of zero charge (p.z.c.) of the particles, whereas rapid coagulation is promoted in the vicinity of the p.z.c. (Sposito, 2004). In many highly-weathered, tropical forest soils, natural organic matter (NOM) coatings on soil colloids can dominate particle surface charge characteristics, yielding a p.z.c. below pH 4 (Chorover and Sposito, 1995). Dissociation of NOM functional groups, particularly above the pKa of carboxylic acids (ca. 4.5), contributes significantly to the development of negative surface charge and interparticle repulsion, especially if mineral colloids are coated with organic matter (Kretzschmar et al., 1993; Chorover and Sposito, 1995; Kretzschmar et al., 1998; Tombácz et al., 2004). The Fe redox cycle affects soil processes through the stoichiometry of oxidation/reduction reactions as well as by producing or consuming high surface area ferric solids and catalytically reactive ferrous species (Hansel et al., 2003). Thus, the

104

interaction of Fe redox processes with colloid chemistry is complex. Reduction of ferric oxides under suboxic conditions can promote colloidal dispersion by dissolving cementitious solids in aggregates (Tadanier et al., 2005). This mechanism was initially invoked to explain colloid release during the intrusion of anoxic water into an Fe-rich aquifer (Ryan and Gschwend, 1990, 1992). Yet, later experiments suggested that concurrent shifts in aqueous phase parameters (pH and ionic strength) produced similar colloidal dispersion without the dissolution of Fe-oxides (Ryan and Gschwend, 1994; Bunn et al., 2002). Indeed, the same solution chemical parameters that govern colloidal stability in oxic systems—pH, ionic strength and ionic composition (Sposito, 2004)—are strongly influenced by dynamic shifts in Fe redox status. This complexity is further accentuated by the fact that soil microbes catalyze, at minimum, the reduction portion of the Fe redox cycle. Thus, concentrations of electron donors (e.g., organic C) and acceptors (e.g., O2, NO3-, etc.) and microbial population dynamics potentially influence colloidal stability in soils via biogeochemical pathways. It is well known that Fe reduction results in hydroxide production and a corresponding increase in soil solution pH (Gillespie, 1920; Joffe, 1935; Ponnamperuma et al., 1966; Vesparaskas and Faulkner, 2001; Kirk, 2004). Conversely, Fe oxidation induces an equivalent proton production. Despite the known impact of Fe redox chemistry on transient pH shifts, few studies have considered the interaction of redox fluctuations and colloidal dispersion. No other soil process results in pH shifts as large as one or more pH units over time scales as short as days to weeks (Kirk, 2004). Furthermore, alternating redox environments also provide the potential for direct, de novo

105

colloid synthesis. Oxidation of FeII(aq) can result in the neoformation of colloidal Feoxides and/or FeIII–NOM complexes (Pullin and Cabaniss, 2003). Thus, Fe redox shifts can influence colloidal mobilization directly via dissolution and precipitation or indirectly via changes in solution chemistry and particle surface charge. This work focuses on the effects of Fe reduction-oxidation cycles (a complete Fe redox cycle is defined here to comprise the reduction of FeIII to FeII followed by its reoxidation back to FeIII) on dissolved and colloid-bound NOM and metals in a soil system. The research objective is to assess the influence of multiple Fe redox oscillations on the mass and chemical composition of suspended colloids, with a particular focus on implications for mobilization of colloid-bound “refractory” elements. We postulated that there would be oscillatory—as well as long-term or cumulative—effects of multiple redox oscillations on colloid chemistry. Specifically, we hypothesized that oscillation in Fe redox status would (1) trigger in situ colloid mobilization during the reducing cycles and (2) promote a long-term increase in colloid stability. Our approach was to force systematic redox oscillations on stirred soil suspensions while monitoring the elemental composition of various particle-size fractions. Our results indicate that colloid-bound refractory elements are mobilized during Fe reduction events, suggesting these elements can not be considered as “immobile” in redox dynamic environments.

106

2. MATERIALS AND METHODS 2.1 Overall approach

Our experiments were designed to model the conditions of redox dynamic environments, specifically upland forest soils subjected to humid, tropical climate. This study also included a detailed mineralogical assessment (Thompson et al., Accepted) and additional experimental details are discussed therein. We targeted Fe valence state transitions by using a “soil-Fe”:FeII equilibrium model to define progressive Eh-triggered mixtures of air or N2 (g) during the oxidizing half-cycles and real-time DOC measurements to define C additions during the reducing half-cycles. 2.2 Soil Selection

A soil was selected from a within a well-characterized climate gradient on the island of Maui, HI that has documented Eh fluctuations (Miller et al., 2001; Schuur and Matson, 2001). The soil contains short-range ordered (SRO) mineral-organic complexes that are characteristic of intermediate (400 ka) weathering stage in basalt-derived soils subjected to a humid climate (Chorover et al., 2004). Details on the sample collection, handling and mineralogical characterization can be found in Thompson et al., (Accepted). A summary of the soil chemical composition is given in Table 1. 2.3 Redox reactor design

A redox reactor system was designed and constructed to control Eh and pH status of triplicate soil suspensions (Thompson et al., 2006). Redox conditions are controlled by automatic modulation of the constant gas flow between ultra-pure air and N2 to maintain selected Eh values using a set of three ORP controllers (Oakton Instruments, Vernon

107

Hills, IL USA). All gases were delivered to the reactors through a 4 M NaOH CO2 trap/humidifier. 2.4 Redox oscillation experiment

2.4.1 Reactor operation Soil suspension incubations were conducted at 25°C, in the absence of light using field moist soil added to the reactors at a solid (dry mass equivalent) to solution mass ratio of approximately 1:11 in an initial 0.002 M NaCl background electrolyte solution. The precise solid concentration was determined at each sampling point by oven-drying (110°C) a pre-measured mass of the slurry for 24 h prior to re-weighing. Several external constraints were applied to isolate the cumulative effects of multiple redox cycles. (1) A full oscillation consisted of 7 d of Fe reducing conditions followed by 7 d of Fe oxidizing conditions. (2) On the second day of each oxidizing halfcycle the pH was reset to 4.5 using 4 M NaOH or 0.5 M HCl. This measure was necessary to curtail the acidifying effects of organic acid accumulation that affected Fe transformation rates with increasing time in preliminary experiments. (3) The rate of O2 (g) addition during the oxidizing half-cycles was controlled by setting a target Eh that would produce an FeII (aq) concentration of 1 µM in equilibrium with “soil Fe” [Fe(OH)3, assuming Kdiss = 15.74; Lindsay, (1979)]. This required adjusting the Eh set-point as the pH changed during the oxidizing half-cycle. (4) A supplemental C source (sucrose) was added to the reactors at the beginning of each Fe reducing half-cycle (excluding the first half-cycle) to give a consistent initial aqueous C concentration of 100 mg L-1. All reactors were operated for 8 wk except reactor C, which malfunctioned in the middle of week 7.

108

2.4.2 Redox sampling scheme The reactors were sampled on days 1, 4 and 7 of each half-cycle. The sampling scheme included taking four 1-mL samples for targeted centrifugal separation of colloidal and dissolved suspension constituents in < 160 nm, < 30 nm, and < 3 kDa particle size fractions. An additional 1 mL sample was collected for immediate tracking of FeII and DOC concentrations (< 30 nm) and one 0.5 mL sample was collected for dry weight analysis. All sample vials were flushed with N2 gas immediately before and after adding the sample. Each sealed sample was transferred to the 95%:5% N2:H2 glove box, and then opened and allowed to equilibrate for 5 min before resealing. After this point samples were opened only inside the glove box. Samples were removed from the glove box and particle size-fractionated using differential centrifugation in an Eppendorf 5417C centrifuge with an F 45-30-11 rotor into < 160 nm (3 min, 7,200 RCF) and < 30 nm (30 min, 18,000 RCF) size classes. Particle size fractions were estimated from Stokes’ law assuming spherical geometry and a mean density of 2.72 g cm-3. Mean particle density was calculated from a preliminary elemental analysis of the colloids (Table 1) assuming Al as gibbsite, Fe as goethite, Ti as anatase, Si as quartz and C as organic matter containing 50% C by mass. After separation, samples were returned to the glove box where the supernatant solutions were removed and acidified to pH 1 by addition of 6 M trace metal grade HCl (0.8% of solution volume). The < 3kDa fraction was centrifuged identically to the < 30nm fraction, then passed through a 3 kDa molecular weight cutoff (MWCO) Millipore AmiconMicrocon YM-3 regenerated cellulose filter using an Eppendorf Minispin Plus centrifuge.

109

The YM-3 filter apparatus does not seal tight enough to prevent the intrusion of O2, thus based on our preliminary experiments this portion of the particle separation was conducted inside the glove box. 2.5 Batch pH shift experiment

To measure the effects of pH alone, a separate set of experiment was conducted under fixed oxic conditions. Batch soil suspensions were equilibrated at 0.025 and 0.05 M ionic strength and pH 4 to 5.5, spanning the range observed in the redox experiments. Preparation (sieving, etc.) and mass concentration of soil solids were the same as described in the redox experiments. The sieved soil slurry was removed from the glove box and bubbled with humidified air overnight to create an “oxic” sample set. The suspension was divided into sixteen 30 mL polypropylene copolymer (PPCO) centrifuge tubes and measured aliquots of 0.5 M KOH were added to give duplicate systems at pH 4.0, 4.5, 5.0 and 5.5 at each ionic strength. The samples were placed on a 7-rpm endover-end rotator for 2 h, and then sampled in a manner identical to that described for the redox samples (e.g., differential centrifugation). 2.6 Chemical analysis

Redox reactor pH and Eh voltages were logged continuously using a custom designed voltage amplifier/processor that fed a signal to a data logger (B&B Electronics, Ottawa, IL). Gel-filled pH (VWR scientific products) and Eh (Omega, Stamford, CT) electrodes were used because of low electrolyte leak rates. All suspension solutions were prepared similarly for analysis of FeII by UV-Vis spectrophotometer (Shimadzu UV-3101 PC) using the Ferrozine method (Stookey, 1970). Interferences in the measurement of

110

FeII from colloidal material or FeIII-ferrozine complexes were identified by comparing UV absorbance at 562 nm and 500 nm and removed by acidification and centrifugation as necessary (Thompson et al., Accepted). Total metal concentrations (Fe, Si, Al, Ti, K, Mg, Ca, Na, Zr, Nb, La, and U) were measured by inductively coupled plasma mass spectrometry (ICP-MS, Perkin Elmer DRC II, Wellesley, MA). Organic carbon (TOC) in the aqueous and solid phases was analyzed by high temperature combustion and subsequent infrared detection of CO2 on a Shimadzu TOC V-CSH analyzer equipped with both solution phase and solid phase (5000A-SSM) sample modules. Glycerin contamination of the Amicon-Microcon YM-3 regenerated cellulose filters prevented accurate measurement of C in the < 3kDa particle size fraction. The filters are saturated with glycerin by the manufacturer to prolong their shelf life and, despite several attempts to remove it with successive acid and water leachings, the number of washes required invariably resulted in destruction of the filter membrane. All concentration values are expressed in units of moles per kilogram of dry soil to compensate for any small variations in suspension solid concentration from sample to sample. During the redox experiment, some soil particles were deposited onto the upper portion of the reactor vessel, effectively removing them from the bulk suspension, and resulting in a minor (< 4%) increase in the suspension water content over the course of the experiment. Suspension dry mass was determined each time a sample was taken. Based on quality control checks on ICP-MS standards, which display an analytical error between 5 – 10% depending on the analyte, we conservatively estimate

111

all ICP-MS analytical error at 10% for individual samples. Error bars are not included on the figures because in many cases they would be smaller than the symbol size, and because this is an estimated error based on repeated analysis of an external standard. The error between replicate reactors is much larger than the analytical error and is the best display of data uncertainty. 2.7 TEM preparation and analysis

Soil colloids were isolated for direct examination by transmission electron microscopy (TEM) by equilibrating a soil suspension at 0.025 M ionic strength (NaCl) and pH 6, size fractionating to obtain a suspension of particles less than 160 nm in size (see section 2.4.2), and then preparing this suspension for TEM examination following method of horizontal centrifugation described in Perret et al. (1991). In a Teflon tube, 50 µL of the colloidal suspension was mixed with 10 µL of medium hardness hydrophilic Nanoplast resin (SPI supplies, West Chester, PA). Five microliters of this mixture were added to a TEM grid (200 mesh Cu, collodion and carbon coated; SPI supplies) and the grids placed on a piece of silicon gel that was affixed to exact center of a microcentrifuge lid with double-sided tape. The grids were spun for 10 s at 7000 rpm, placed on a Teflon block, dried for 24 h under a laminar flow hood at room temperature, and then dried at 50˚C for 12 h. Parallel blanks were prepared each time a sample was processed to test for particulate contamination. TEM micrographs were obtained using a Hitachi H-8100 operating a 200 KeV with a point to point resolution of 2.3Ǻ. The chemical composition of individual particles was determined with a Thermo Noran digital energy dispersive spectroscopy (EDS)

112

system with an accelerating voltage of 20 KeV, a live time of 100 s, and a take of angle of 30°. 2.8 Notation

Element concentrations are denoted throughout the text as [M]